Low bit rate voice transmission for use in a noisy environment

• US 5,353,374 A
• Filed: 10/19/1992
• Issued: 10/04/1994
• Est. Priority Date: 10/19/1992
• Status: Expired due to Term

• Current Assignee: Lockheed Martin Corporation (Martin Marietta Corporation)
• Original Assignee: Loral Aerospace Corp. (MHR Holdings LLC (New York))
• Inventors: Wilson, Dennis L., Wayman, James L.
• Primary Examiner(s): Knepper, David D.
• Application Number:
  US07/963,101
• Time in Patent Office: 715 Days
• Field of Search: 381/29-40, 395/2.35, 395/2.28, 395/2.25-2.27, 395/2.36, 395/2.37
• US Class Current: 704/226
• CPC Class Codes: G06T 9/004 Predictors, e.g. intraframe, interframe coding
• CPC Class Codes: G06T 9/005 Statistical coding, e.g. Huffman, run length coding
• CPC Class Codes: G10L 19/04 using predictive techniques
• CPC Class Codes: G10L 21/0364 for improving intelligibility

Abstract

A compressing a voice signal by the steps of (a) digitizing an input signal that includes a voice signal, the input signal including a coherent noise component; and (b) compressing the digitized voice signal with a synchronized overlap add processor (20). So as to prevent the synchronized overlap add processor from locking to the coherent noise component, the step of compressing includes an initial step of applying the digitized input signal to a linear predictor (16), the linear predictor having time constants selected for attenuating the coherent noise component of the input signal. The residual signal output of the linear predictor includes the voice signal, and an uncorrelated noise component if one is present in the input signal. The operation of the synchronized overlap add processor also functions to attenuate the incoherent noise component. Further compression of the compressed voice signal is accomplished by Huffman coding, arithmetic coding, or transform coding so as to provide a greatly compressed voice signal that, when subsequently expanded, is found to exhibit excellent voice quality.

Claims

1. A method for processing an audio signal, comprising the steps of:

digitizing an input signal that includes an audio signal component, a coherent noise signal component, and a non-coherent noise component;

applying the digitized audio signal to a linear predictor means, the linear predictor means having a plurality of serially coupled time delay elements wherein a time delay between adjacent time delay elements is selected for attenuating the coherent noise component of the input signal, said linear predictor means having a residual output signal that is representative of the audio signal component and the non-coherent noise component, less the coherent noise component;

applying the residual output signal to a synchronized overlap add processor means for compressing the audio signal component in accordance with a compression ratio, the synchronized overlap add processor means furthermore attenuating the non-coherent noise component of the input signal; and

further compressing the audio signal component with encoder means to provide a digital output signal that is expressive of the audio signal component.

2. A method as set forth in claim 1 wherein the compression ratio is within a range of approximately 4:

1 to approximately 16;

1.

3. A method as set forth in claim 1 wherein the step of further compressing is accomplished by a coding method selected from the group consisting essentially of a Huffman coding method, an arithmetic coding method, and a transform coding method.

4. A method as set forth in claim 1 wherein the step of further compressing includes an initial step of companding the compressed audio signal component.

5. A method as set forth in claim 1 wherein the step of applying the residual output signal having the attenuated coherent noise component to a synchronized overlap add processor means includes an initial step of first applying the residual output signal having the attenuated noise component to a differential processing means that includes means for delaying a sample of the audio signal component, and for subtracting the delayed sample from a current sample of the audio signal component.

6. A method for compressing a voice signal, comprising the steps of:

digitizing an input signal that includes a voice signal component to provide a digitized input signal, the digitized input signal further including a coherent noise component; and

compressing the digitized input signal with a synchronized overlap add processor means, wherein, so as to prevent the synchronized overlap add processor means from locking to the coherent noise component, the step of compressing includes an initial step of

applying the digitized input signal to a linear predictor means, the linear predictor means having a plurality of serially coupled time delay elements wherein a time delay between adjacent time delay elements is selected for attenuating the coherent noise component, said linear predictor means having a residual output signal that is representative of the voice signal component, less the coherent noise component, the residual output signal being applied to an input of the synchronized overlap add processor means for compression thereby.

7. A method as set forth in claim 6 wherein the digitized input signal further includes a non-coherent noise component, and wherein the step of compressing includes a step of attenuating the non-coherent noise component of the digitized input signal.

8. A method as set forth in claim 6 and further including a step of further compressing the digitized input signal by encoding the compressed digital voice signal component to provide a digital output signal that is expressive of the voice signal component, wherein the step of encoding is selected from a coding technique selected from the group consisting essentially of a Huffman coding technique, an arithmetic coding technique, and a transform coding technique.

9. A method as set forth in claim 6 wherein the step of compressing compresses the digitized voice signal component in accordance with a compression ratio in the range of approximately 4:

1 to approximately 16;

1.

10. A method as set forth in claim 6 wherein the step of applying is followed by a step of applying the residual output of the linear predictor means to a differential processing means that includes means for delaying a sample of the digitized voice signal component, and for subtracting the delayed sample from a current sample of the digitized voice signal component.

11. A method as set forth in claim 6 and further comprising the steps of:

further compressing the digitized voice signal component by encoding the compressed voice signal component in accordance with a technique selected to provide a digital output signal that is expressive of the voice signal component;

wherein the step of applying is followed by a step of applying the digitized voice signal component having the attenuated coherent noise component to a first differential processing means that includes means for delaying a sample of the digitized voice signal component, and for subtracting the delayed sample from a current sample of the digitized voice signal component; and

a step of

expanding the further compressed digitized voice signal component, the step of expanding including the steps of,

decoding the further compressed digitized voice signal component in accordance with the technique selected to provide a partially expanded voice signal component;

decompressing the partially expanded digitized voice signal component with a second synchronized overlap add processor means;

applying the further decompressed digitized voice signal component to a second differential processing means that includes means for delaying a sample of the decompressed digitized voice signal component, and for adding the delayed sample to a current sample of the decompressed digitized voice signal component; and

converting the decompressed digitized voice signal component to an analog voice signal.

12. Apparatus for compressing a voice signal, comprising:

means for digitizing an input signal that includes a voice signal component, the input signal further including a coherent noise component; and

means for compressing the digitized input signal, said compressing means including a synchronized overlap add processor means, wherein, so as to prevent the synchronized overlap add processor means from locking to the coherent noise component, the apparatus further includes

linear predictor means interposed between said digitizing means and said compressing means, said linear predictor means having a plurality of serially coupled time delay elements wherein a time delay between adjacent time delay elements is selected for attenuating the coherent noise component, said linear predictor means having a residual output signal that is representative of the voice signal component, less the coherent noise component, the residual output signal being applied to an input of the synchronized overlap add processor means for compression thereby.

13. Apparatus as set forth in claim 12 wherein the input signal further includes a non-coherent noise component, and wherein said synchronized overlap add processor means includes means for attenuating the non-coherent noise component of the input signal.

14. Apparatus as set forth in claim 12 and further including encoder means for further compressing the compressed digitized voice signal component to provide a digital output signal that is expressive of the voice signal component, wherein the encoder means is selected from a group consisting essentially of a Huffman coding means, an arithmetic coding means, and a transform coding means.

15. Apparatus as set forth in claim 12 wherein the compressing means compresses the digitized voice signal component in accordance with a compression ratio in the range of approximately 4:

1 to approximately 16;

1.

16. Apparatus as set forth in claim 12 and further including differential processing means interposed between said linear predictor means and said compressing means, said differential processing means including means for delaying a sample of the digitized voice signal component, and means for subtracting the delayed sample from a current sample.

17. Apparatus as set forth in claim 12 and further comprising:

means for further compressing the digitized voice signal component by encoding the compressed digitized voice signal component in accordance with a technique selected to provide a digital output signal that is expressive of the voice signal component;

first differential processing means interposed between said linear predictor means and said compressing means, said first differential processing means including means for delaying a sample of the digitized voice signal component, and means for subtracting the delayed sample from a current sample; and

further including means for expanding the further compressed digitized voice signal component, said expanding means including,

means for decoding the further compressed digitized voice signal component in accordance with the selected technique to provide a partially expanded digitized voice signal component;

means for decompressing the partially expanded digitized voice signal component with a second synchronized overlap add processor means;

second differential processing means that includes means for delaying a sample of the decompressed digitized voice signal component, and for adding the delayed sample to a current sample of the decompressed digitized voice signal component; and

means for converting the decompressed digitized voice signal component to an analog voice signal.